In 2009, a usability researcher named Jared Spool wrote up one of the most consequential case studies in the history of UX design. He didn't name the brand publicly — his original article on UIE (User Interface Engineering) called it "a major retailer" — but anyone who's worked in e-commerce knows the story. The fix was a single button. The revenue impact was estimated at $300 million in the first year.
Here's what happened.
The retailer in question had a standard e-commerce checkout. Add to cart, then a screen with two options: Log In or Register. New customers were forced to create an account before they could complete the purchase. The product team's logic was reasonable on paper — once a customer registered, future purchases would be faster, and the company could collect data for retention marketing.
Spool's team ran user research on the flow. What they found was unprintable in most corporate decks. Users hated being forced to register. The single most-quoted line from the research transcripts was a customer saying, "I'm not here to be in a relationship. I just want to buy something."
Spool's recommendation was almost stupidly simple. Add a third button: Continue as Guest. Don't require registration. Let people buy and leave if that's what they want.
The team A/B-tested it. Conversion rose by 45%. Annualized revenue impact was estimated at over $300 million.
That single button is, in a sense, the cleanest example of friction as a strategic variable in the history of conversion optimization.
What "Friction" Actually Means in Behavioral Terms
Friction, in the behavioral economics sense, is anything that increases the cognitive or physical cost of completing an action. Donald Norman's The Design of Everyday Things — still the foundational text on how design shapes behavior — makes the case that every step in a process is a potential leak point. Steve Krug's Don't Make Me Think extends the argument to digital interfaces: every micro-decision you ask a user to make is a small tax on their willingness to continue.
The $300M button story is a friction-removal story. The forced-registration step was a cognitive ask — think about your future identity with this brand — at a moment when the user was in transaction mode. The two mental contexts didn't fit together. The user's brain rebelled. The cart was abandoned.
If you've read Ron Kohavi, Diane Tang, and Ya Xu's Trustworthy Online Controlled Experiments — the closest thing the A/B-testing field has to a manual — you'll see this exact pattern documented across hundreds of experiments at major tech companies. Roughly two-thirds of A/B tests fail to move the metric. The ones that win disproportionately remove steps rather than add them. The compounding wisdom of two decades of large-scale experimentation can be summarized in one sentence: subtraction beats addition almost every time.
When Friction Is Good
The interesting nuance is that not all friction is bad. The same mechanism that costs you $300M when applied to the wrong moment can create value when applied to the right moment.
A few categories where friction works in the brand's favor:
Status goods. The Hermès Birkin waitlist is a feature, not a bug. The difficulty of acquisition is part of the product's value. I've written about this elsewhere as the Veblen Goods phenomenon. Stripping the friction would destroy the brand.
High-stakes confirmations. When the user is about to do something irreversible — delete an account, transfer a large sum, send a sensitive email — adding friction is a service. The two-step confirmation Slack uses for deleting messages, the GitHub "type the repository name to confirm deletion" pattern, the AWS console's deliberately ugly delete flows. All of these are friction-as-protection.
Identity signals. Slow ketchup. Hard-to-find shortcuts. The deliberate "we sell out fast" framing of streetwear drops. Each of these uses friction to signal value or scarcity.
The diagnostic question, when you're looking at any friction point in your own flow, is: does this step create value for the customer, or only for the business? If it's the second, remove it. The math will look like the $300M button.
What I Take From This
The most useful instinct I've developed from years of looking at conversion funnels is removal bias — the discipline of always asking "what can I take out?" before asking "what can I add?"
Most teams have the opposite reflex. Adding features feels productive. Removing features feels like admitting failure. The economic incentive in most product orgs is to ship more, not less. Spool's $300M button is a quiet reminder that the most expensive UX decisions are sometimes the ones you make by not removing the thing that's already there.